Another question which has always been on my mind is why is there that obnoxious dust trap on the back of many refrigerators? ( .. and where did it go on the sleek new ones?) I had heard growing up that it was supposed to be the heat exchanger that allowed the magic stuff in the refrigerator to get rid of the excess heat inside. This seemed pretty reasonable at the time since it was usually warm back there. As I grew older I came to wonder why there had to be all those little wires all over the purported heat exchanger. They were so small that there couldn't be anything flowing through them to give up heat so it didn't make any sense. It was with these thoughts in mind that I set out to try and shed some light on this question a couple of years ago.

The first step in this process was to clarify what I thought was actually happening in the heat exchanger. Well, that seemed pretty straight forward - it is to get rid of heat from inside the frige! I guess it does that by doing some magic which takes heat from the milk inside and puts it into a fluid which is circulated through a tube to the exchanger. In the heat exchanger the fluid must be hotter that the surrounding air and loses its heat through the metal to the air. My previous experience suggested that the metal tube could really conduct a lot of heat to the air in a very short time. It didn't seem reasonable that there needed to be so much metal on the heat exchanger. And that, as they say, must be the problem. I needed to get a clearer idea of how metals transported heat into the air. As a result I set up this lab to make some measurements of heat flow through a variety of metal cans.

The first thing I needed was a selection of metal cans. After finding several sizes I then realized I needed to know quantitatively how much heat was moving through the metal. I noted that I could exploit the known properties of water for this purpose. The thermal property I exploited is known as the heat capacity of the material. This is an property with which you are all familiar that says that if I put heat into some object it gets warmer (amazing!). How much warmer it gets depends dramatically on the material. Wood not only conducts heat slowly it heats up slowly as well. Metals tend to conduct heat well and they also heat up pretty quickly. Now water is a handy material which has been extensively studied and its specific heat capacity is well understood. This property of water can be summarized as "4.18 J of energy (heat) will change the temperature of 1 gram of water by 1 degree Centigrade (or Kelvin if you prefer)". This meant that by measuring the change in the temperature of the water inside my cans I could determine how much heat had left the water through the can. If I wanted to I could also calculate the thermal conductivity of the metal. Much to my dismay, when I tried to do this I got ridiculous numbers for the thermal conductivity of my metal cans. The numbers were way too low! After some thought I realized that there was another important process taking place that I had not taken into account. The metal was conducting the heat from the water to the air but until the air carried the heat away the water wouldn't cool down. Air picks up heat by conduction but carries it away via convection. This is a slow and not very efficient process.

After some more thought I decided that it was OK. What I was measuring on the lab bench was the same thing that was taking place on the refrigerator. What I really needed to know now was how much heat could the air pick up from the metal in some period of time. Because the process of removing the heat from the water, passing it through the metal can, and having the air pick up that heat is inherently a mixed bag of several processes we can't easily apply our basic physics tools. In these circumstances science often seeks to define a number which seeks to describe a complex process by it's apparently more simple behavior. This process is a form of complex thermal conduction which we could describe by our thermal conductivity equation if we had an appropriate number for the coefficient of thermal conductivity. Even though we know it is not the same thing we had "characterize" the process as thermal conductivity and determine an "effective coefficient of thermal conductivity - k_eff - for this process.

With this in mind I went back to work and reassesed my numbers. To my suprise there was not a lot of difference in the results from the different cans. In the end I had a handy number - k_eff - (which I have long since forgotten) that told me how much heat the air could transport away from each square meter of metal (at a temperature of roughly 50 degrees Centigrade) in a second. This number mixes together the thermal conductivity of the water, the metal can, and the convection of the air into a single quantity and ignores the thickness. Now I was ready to consider the refrigerator again.

It was becoming apparent that if you were patient enough even a very small heat exchanger should be able to cool down a refrigerator. The question then becomes "how long do I want to wait?". So I had to decide what a reasonable task for a refrigerator ought to be. I chose some water-like items to put in my imaginary refrigerator and allowed a plausible time for them to cool down. From this I was able to calculate how much heat had to be removed from the refrigerator in a defined period of time. This same amount of heat had to be carried away from the heat exchanger by the air. Some quick measurements immediately showed why the wires had to be there. I hope that you will feel the same sense of satisfaction that I felt when I worked this out. I have been intentionally vague about many parts of the actual experiments because I want you to experience the process as much as possible. Good Luck and I'll see you in Lab!!!

A refrigerator is an object that is intended to move energy (heat) from a cold place to a warm place. A heat pump has an identical mission in life. A heat pump ideally takes heat from the already cold out of doors and moves it into your house. So why do they look so different if they do the same thing? The principle difference is that the cold part of a refrigerator is a small space inside while the cold part of a heat pump is the whole outdoors. What this means is that I should be able to take a refrigerator and with suitable modifications use it as a heat pump to warm my house. We may also explore this odd idea in this lab.

Reading:

Read about specific heat (or sometimes specific heat capacity which is ever so slightly different than heat capacity)

Know the formula and particularly how it applies to heating and cooling water.

Here's a link to Wolverine Tube Inc which is a company which worries about these same sorts of heat transfer problems. Their data book has loads of tech info and perhaps explains why we are using the simple model we are for this lab. I haven't explored it much but it seems very interesting.